Early detection of myocardial iron deposition is mandatory to prevent irreversible heart failure, but systolic function is often preserved in iron induced cardiomyopathy. The assessment of diastolic function using CMR and its relation with myocardial iron content determined by myocardial R2* measurements has not been sufficiently investigated. The purpose of this study was to investigate the relation of diastolic function and myocardial iron overload in patients with systemic iron overload disease using cardiac magnetic resonance imaging (CMR) with different volumetric approaches.CMR was performed in 124 patients (67 female/57 male, mean age 29.3±14.3y) with the diagnosis of transfusion dependent thalassemia (TDT: 37f/33m), Diamond-Blackfan anemia (DBA: 10f/5m), sideroblastic anemia (SBA: 11f/9m), Friedreich ataxia (FA: 1f/4m), hereditary hemochromatosis (HHC: 8f/6m) and 23 healthy controls (6f/17m). Beyond TDT, most patients with DBA and SBA had received regular chelation therapy and blood transfusions. None of the volunteers had a history of cardiac disease. CMR protocol CMR was performed using a four-element phased array coil on a 1.5T imager and performing single slice (2D) acquisitions with breath hold retrospective ECG gated gradient recalled echo (GRE) sequences. Phase encoding steps (segments per view) were reduced for actually increasing heart rates. Specifically, cardiac function was assessed from short axis cine-series (25 phases). Data were acquired by a SSFP sequence with the following parameters: TE=1.6ms, readout TR=50ms, FA=65°, bandwidth=965Hz/pixel, slice thickness=6 mm. For the assessment of the transverse relaxation rate R2*, data were acquired in a mid-papillary short axis slice (10mm) by a breath hold retrograde ECG-gated sequence with 9 heartbeats in end-diastole and a readout time (TR) of 244ms using 12 bipolar echoes per breath hold (TE= 1.3-25.7, Δt=1.16ms), FA=20°, and bandwidth=1955Hz/pixel. LV end-diastolic volumes (EDV), end-systolic volumes (ESV), EF and myocardial mass (M) were determined by manual delineation of endocardial and epicardial borders in end-systolic, end-diastolic and mid-diastolic short axis views. LV blood and myocardial volumes can then be calculated for every of the acquired 25 phases of the cardiac cycle. LV volumetry was assessed in two different ways 1) including trabeculae and papillary muscles (TPM) into the LV cavity (TPM-in) by manual definition of the myocardial borders and 2) excluding TPM from the LV cavity (TPM-ex) by adjustment of the individual tissue-blood threshold of the relative color contrast scale (Figure1). The calculated volume time curve data were exported to EXCEL. The temporally differentiated LV volume curve usually results in 3 peaks characterized by the systolic peak contraction rate (PCR) and the diastolic early and atrial peak filling rate (EPFR and APFR). The EPFR and APFR assessed by CMR reflect the early (E) and atrial (A) transmitral peak filling velocities determined by echocardiography. In correspondence to echocardiography (E/A ratio), the peak filling rate ratio (PFRR = EPFR/APFR) is the equivalent to characterize diastolic filling patterns using CMR and is associated with cardiac iron content (1). Intra- and inter-operator variability was tested for experienced and inexperienced operators by an Altman-Bland plot. Further parameters that were determined included heart rate (HR) during MR examination, the body surface area (BSA), and cardiac output (CO). Contractility (CTY) was determined noninvasively according to Zhong et al. (2) as CTY = 1.5·PCR/MV with MV=LV mass volume= M/1.05, which refers to the maximum rate of change of intracavity pressure-normalized wall stress. The transverse relaxation rate R2* was determined in a mid-papillary short axis slice from region of interest (ROI) based signal intensities of the cardiac septum. A mono-exponential function of TE with a signal level offset was fitted to the signal intensities using the Levenberg-Marquardt algorithm (3). The in vivo liver iron concentration (LIC: dry-weight conversion factor = 6) was measured by biomagnetic liver susceptometry (4). Nonparametric statistics was applied, especially for R2*, EDV, ESV, EF, and PFRR: median, 95% range, (Wilcoxon-) Mann-Whitney and Spearman rank correlation (rS) test. For adjustment (prediction) of LV function parameters, especially EF and PFRR, by age, HR, and R2*, a parametric univariate and multivariate regression analysis was used after logarithmic transformation of R2*, which resulted in coefficients of determination (r2), regression coefficients m, standardized regression coefficients B (Beta), and their contributing significance p. ROC analysis was performed.Only the TPM-ex method allowed accurate determination of the EPFR, APFR and the PFRR, respectively (Figure 2). The exclusion of TPM resulted in significant relatively decreased EDV (17%), ESV (36%) and increased EF (11.6%, p<0.0001) compared with values by inclusion of TPM. In 5 patients, EF increased beyond the 95% limits of agreement of the underlying Altman-Bland plot (-3.9% to 27%). The peaks in the differentiated temporal stroke volume representing the PCR, EPFR, and APFR could be determined with sufficient precision only by excluding TPM (Figure 2). The relative cardiac tissue-blood threshold, which is proportional to the amount of TPM, ranged from 14% to 41%. Significant correlations were found with R2* (p=0.007), marginally with EF (p=0.042), but not with the cardiac mass (p=0.6). Healthy controls (R2*=21-38s-1) and patients (R2*=25-430s-1) revealed LV ejection fractions (EF) of 63-77% and 10-84%, respectively. The PFRR ranged from 0.9-4.1 in controls and 0.9-6.8 in patients. Multivariate regression analysis predicted age, heart rate, and log(R2*) to be the only equivalently significant predictors of PFRR (r2=0.37, p<10-4). ROC analysis revealed increased sensitivity to indicate myocardial iron overload for the PFRR (AUC=0.75, p<10-4) compared to EF (AUC 0.59, p=0.03) (Figure3). In healthy controls, the 95%-range of septal R2* was 21-38 s-1 (maximum rate = 37.7 s-1). Median R2* rates were significantly increased in patients with TDT (p<10-4), DBA (p=0.001), and SBA (p=0.002), but within the normal range in patients with HHC and FA. EF did not differ between patients and controls. Contractility was reduced only in FA patients (p=0.007) due to a higher cardiac mass index (p=0.04) and tighter range. The diastolic PFRR was significantly increased in TDT (p=0.004) and SBA (p=0.03) patients, marginally in DBA (p=0.06), but decreased in HHC (p=0.002). ROC analysis for septal cardiac R2* > 40 s-1 revealed an area of 0.63±0.05 (p=0.005) for EF and 0.76±0.04 (p<10-4) for PFRR, with equal sensitivity and specificity rates of 61% and 69%, at corresponding cut-off levels of EF = 68% and PFRR = 2.49. It is also shown that EF values > 69% (no discrimination from unity line) have no further significance for cardiac iron loading. Using the 95% range limits of our control group as cut-off levels for, sensitivities and specificities of 45% and 91% were obtained for EF < 63%, and for PFRR > 4.1 we obtained 25% and 96%, respectively.The diastolic peak filling rate ratio (PFRR) is a sensitive marker to indicate diastolic dysfunction from myocardial iron toxicity in patients with systemic iron overload disease. Precise assessment of the PFRR by CMR requires a volumetric approach with exclusion of trabeculae and papillary muscles from the LV cavity. The PFRR assessed by CMR may be a valuable parameter for the screening and monitoring of myocardial iron toxicity due to iron deposition in patients with preserved systolic function.